I can’t wait until I can get in the driver’s seat of a car, input my destination into the GPS, and then take a nap as my car drives me there. This sci-fi future, I believe, is closer than many people think.
Americans won’t just accept a fully autonomous self-driving car from out of nowhere. It will have to happen piece by piece, where AI takes over one small thing then the next. And indeed, this is what we find happening.
In 2006, Lexus introduced a car that could parallel park itself. That feature is becoming more and more common. Intelligent braking if the driver is not acting is another recent example of cars driving themselves. This feature has been extolled by the Insurance Institute for Highway Safety. In a recently released report, IIHS announce that this intelligent braking outright prevents 25% of low-speed crashes in the vehicles in which it is installed.
It’s better that computers will eventually drive. Humans just aren’t good at it. The World Health Organization estimates that around 1.2 million people are killed in car crashes each year with around 45,000 of those deaths taking place in the US. Almost all of these are due to human error, something that will be eliminated with self-driving cars.
But how far away are cars that completely drive themselves? Well, I can’t go out and buy one today. But check out this TED talk.
Self-driving cars are already here! Google’s self-driving cars have logged 140,000 miles across cities and highways through both day and night without an accident. And if you think that Google accomplished this feat by relying on their cars driving meekly and 20 mph less than the speed limit, here’s one of their cars screeching tires and hauling ass through an obstacle course.
Once cars are completely driving themselves, even more opportunities open up. Scientists in Italy are working on software that allows cars to talk to each other, with the end goal of connecting every single car on the road. What would this system be used for?
When a car in an accident experiences a sudden change in acceleration, this change would be captured by the sensor and alert cars and drivers approaching the same spot.
In addition to preventing accidents in the first place, self-driving cars will better react to accidents when they do occur, preventing even more accidents. This double whammy, I believe, will have an incredibly dramatic effect on the number of deaths and injuries from auto accidents each year. Once self-driving cars are widespread a 95% reduction of traffic deaths is within the realm of possibility.
One last bit of good news: traffic will also be lessened once self-driving cars are the norm. Every time I sit in a traffic jam, I dream that self-driving cars are already here. There aren’t any idiots switching lanes to try to get ahead only to end up slowing everyone down. There are no rubberneckers braking to stare at the gore of an auto accident. There is no one suddenly accelerating and then stopping, disrupting the flow of traffic.
Instead, there are thousands of self-driving cars proceeding at a quick and steady clip on the highway. People now get to their destination much sooner and rush hours have been greatly alleviated, in some places entirely eliminated. Self-driving cars are following each other closer than would be possible if driven by humans. No one brakes for no reason. There is no road rage. When a two-lane highway merges into one lane, the cars evenly space themselves and merge while barely losing any speed, like a zipper being zipped. Everything is beautiful. Nothing hurts.
So how will this paradise come about? I expect cabs and buses might be the first early adopters, for economic reasons. Maintaining a mass-produced AI will be cheaper than paying individual drivers. There would be some differences, sure. Video surveillance would have to be used to prevent and prosecute hooligans. But in the end, we’d be much better off with self-driving cars. And I believe (or at least want to believe) that I will ride in my first self-driving car before 2020.
No comments:
Post a Comment